(with valuable input from lucien cremaldi, u. miss. simon ... documents... · wirebonds and >25...

Arndt & Shipsey - PurdueCMS SLHC Upgrade Workshop

21-22 May 20081

CMS Upgrade Proposal 7.15

Proposal for US CMS Pixel Mechanics R&D at Purdue and Fermilab in FY08Daniela Bortoletto, Simon Kwan, Petra Merkel, Ian Shipsey (Physicists)

Kirk Arndt, Gino Bolla, Joe Howell, C.M. Lei (Engineers)

Fall, 2007AbstractThis proposal includes R&D of mechanical aspects for a possible 3rd set of disks for the current CMS Forward Pixel detector, a possible intermediate upgrade pixel detector for CMS at the LHC, and the upgrade pixel detector for CMS at the SLHC.

Kirk Arndt, Ian Shipsey – Purdue Univ.(with valuable input from Lucien Cremaldi, U. Miss.

Simon Kwan, CM Lei, Mikhail Kubantsev and Vesna Cuplov, FermilabDaniela Bortoletto, Purdue

on behalf of Purdue and Fermilab pixel mechanics groups)

http://cmsdoc.cern.ch/cms/electronics/html/elec_web/docs/slhcusg/proposals/Pixel_Mech_FY08_RD_request-Purdue-FNAL-v6.pdf


21-22 May 20082

• Goals for upgrade pixel mechanics• Review aspects of the current pixel

detector for improvement• Review what’s in our revised plan• Mechanical R&D that has been started• Revised R&D task list and milestones

Talk Overview


21-22 May 20083

CMS Forward Pixels at Purdue and FNAL• The Purdue group developed the tools,

materials & techniques for assembly, testing and delivery of ~1000 Pixel modules for the CMS FPIX (~250,000 wirebonds and >25 million pixels) at the planned assembly rate of 6 modules per day.

• Rework techniques were also developed at Purdue to recover faulty modules and maximize the final yield.

• The Fermilab group designed, assembled and tested ~250 Panels on 8 Half-Disks (for Pixel module support and cooling), in 4 Half-Cylinders (with cooling and electronics services) for FPIX.

• Fermilab had overall management responsibility for the construction of FPIX, as well as the transportation of detector assemblies to CERN and commissioning of the detector at CERN.


21-22 May 20084

Phase 1 Upgrade Requirements

It is MANDATORY that the Phase 1 UpgradeContribute to the physics by maintaining or improving performance at higher luminosityFit within the available construction and installation timeFit within the available funding

It is also HIGHLY DESIRABLE thatEach part of the Phase 1 Upgrade should be a significant step on the ROADMAP for the Phase 2 Upgrade


21-22 May 20085

Goals for US CMS Pixel Mechanics R&D at Purdue and Fermilab

• In view of the recent Phase 1 upgrade plan, we have revised our mechanics R&D toward a Forward Pixel replacement / upgrade detector in 2013 = 3 disks + CO2 cooling

– Reduce material significantly (and distribute more uniformly)– Reduce # of components and interfaces = simplify assembly– Study alternatives to current disks for detector geometry (i.e. fewer

module types)– Improve routing of cooling, cables, location of control and optical hybrid

boards • A CO2 cooling system may lead to a design that uses significantly

less material, and acts as a “pilot system” for implementation in a Phase 2 full CMS (and ATLAS) tracker upgrade.

• Mechanics R&D compatible with new detector layout and technologies required to maintain or improve tracking performance at higher luminosity + triggering capability

– Serial powering (or other powering scheme)– Longer (possibly thinner) ROC with double buffer size for higher data

rate and HV-cap– MTC (Module Trigger Chip) for pixel-based trigger at Level 1


21-22 May 20086

Current FPIX Material Budget

Most material between 1.2 < η < 2.2 is in Disks

(brazed aluminum cooling loops) and electronics (ZIF

connectors and adapter boards)

and between 2.2 < η < 3.6 in cables and cooling


21-22 May 20087

Present Pixel System with Supply Tubes / Cylinders

10

0

20

20cm 40 60 80 100

FPIX service cylinder

BPIX supply tube

DOH & AOH motherboard+ AOH’s

Power board BPIX end flange printed circuits

η = 1.18

η = 1.54 η = 2.4

Port Cards and AOH

Cooling manifold

Adapter boards


21-22 May 20088

Phase 1 Pixel System Concept• Replace C6F14 with CO2 Cooling• 3 Barrel Layers + 3 Forward Disks (instead of 2)• Pixel integrated modules with long Copper Clad Aluminum pigtail cables • Move OH Boards and Port Cards out

10

0

20

20cm 40 60 80 100

FPIX service cylinder

BPIX supply tube

η = 1.18η = 1.54

η = 2.4

OH Boards+ Port Cards + Cooling Manifold moved out

Long CCA pigtails


21-22 May 20089

Present Mass Distribution in FPIX Service Cylinder

Disks

PortCards and AOH

End Flange

End Electronics

Coils of fibers

Pipes and Cables Back

Pipes and Cables Front

Service Cylinder


21-22 May 200810

Weight and Radiation Length estimates for replacement/upgrade FPIX detector

Present Fpix New FpixMass(g) RL(cm) Mass(g) RL(cm) Comments

Service Cylinder 7212 206 5770 256 reduce CF thickness, most of SC is outside FPIX acceptance

End Flanges 2820 13 2136 27 use CF instead Al, outside FPIX acceptance

End Electronics 2456 71 2456 71 outside FPIX acceptance

Coil Fibers 2271 162 1136 314 use custom (actual) length fibers

Pipes and Cables Back 4860 5.8 1480 29 1mm diam. CO2 pipes, CCA cables

Port Cards and OH 1616 19 1616 19 move out of FPIX acceptance

Pipes and Cables Front 5489 6.3 1647 32 CO2 pipes, CCA cables, move manifold out of FPIX acceptance

Disks 4512 14.7 2345 31 CO2 cooling, CF structure, highTC heat spreaders, half # of HDI

Total 31237 56 18586 103~40% less mass overall

2X less mass in FPIX acceptance (1.5 < η < 2.5) and 2X RL increase


21-22 May 200811

Revised Mechanics R&D Proposal

Most of the tasks listed in our earlier proposal remain parts of our new plan, with renewed emphasis on reducing material significantly.

Our revised plan:– develop a conceptual design early that

integrates CO2 cooling and lightweight support– develop mechanics targeted at the design

(+ revisions)– Goal: optimal mechanical/thermal solution for

the replacement/upgrade FPIX in ~2 years


21-22 May 200812

Revised Mechanics R&D Proposal1. Conceptual design

– Integrate cooling/support structure into an overall detector package and eliminate redundant features

2. Cooling/Support development– Study CO2 cooling, including construction of a CO2 cooling system

for lab bench testing of prototype integrated cooling/support structure and prototype pixel detector integrated modules

• Improved C6F14 is backup cooling solution– Investigate new materials and designs for support/cooling structure

to lower the material budget• Study suitability of composites with high thermal conductivity for

fabrication of low mass support frame and thermal management scheme• Finite Element Analysis of mechanical stability and thermal performance• Composite material combinations (ex: Thermal Pyrolytic Graphite vs. C-F

laminate) for integrated module support • Investigation of alternative cooling channel materials• Design cooling structure in a sparse arrangement that minimizes the

number of fluid connection joints• Measurements of cooling performance of prototype integrated module-

on-support structures, and evaluation of radiation hardness of alternative materials


21-22 May 200813

Revised Mechanics R&D Proposal

3. Integrated Module Development– Evaluation of adhesives for integrated module

assembly and rework– Evaluate state-of-the-art alternatives (ex: ceramic

vs. flex-laminated-on-rigid substrates) for dense multilayer interconnects for readout and power circuits

– Development of (semi-robotic) tooling to assemble prototype integrated modules

– Testing of mechanical, thermal and electrical properties of prototype integrated modules with radiation


21-22 May 200814

What has been started at PurdueIntegrated module development - Adhesives study• Begun a market survey of adhesives for

pixel integrated module assembly that meet requirements for SLHC– Requirements for adhesive:

• Thermal conductivity: > 0.2 W/m-K• Soft: shear modulus < 50 N/mm^2• Radiation hard• Electrically non-conductive• Curing at room temperature• Not flowing during application: adhesive

confined within chip• Good wetting properties• Not creeping after curing• Allow integrated module replacement

without damaging the support• Will build mechanical grade integrated

modules using candidate adhesives and evaluate the mechanical properties after irradiation.

FPIX adhesive sample tensile tested after

irradiation


21-22 May 200815

Integrated module development - Fine pitch wirebond layout

• We can make rows of fine pitch (100 um) bonds with distance vs. height shown in the area bounded by the blue line, but the resulting loop profiles are inconsistent or neighboring wires sometimes touch.

• We can consistently make double (staggered – inner and outer) rows of fine pitch bonds that do not touch each other with distance vs. height shown in the area bounded by the green line.

• At this time, we conclude that we can consistently make double rows of fine pitch step bonds at a height of 2mm with a nominal lateral distance of 3mm.

Wirebond length vs. heightfine pitch (100um), staggered bond pad pattern

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

bond distance (mm)

step

hei

ght (

mm

)

all wirebondsacceptable wirebondsK&S max lengthSVX4 - L00 sensor inner rowSVX4 - L00 sensor outer rowFPix fanout step bondFPix ROC-VHDI nominal step bondFPix ROC-VHDI proposed step bond

neighboring wirebonds touch

irregular wirebond

loops

machine limit constrains loop

height

Wirebond between “longer” ROC and

existing VHDI

bond height

bond length

Recent exploration of parameter space @ Purdue


21-22 May 200816

Current Fpix components stack-up

Currently, FPIX Disks have a lot of material in: – passive Si and Be substrates– flex circuits with Cu traces– thermal conductive (BN powder) adhesive interfaces– brazed aluminum (0.5mm wall thickness) cooling channels


21-22 May 200817

Fpix Blade components and thermal interfaces

• Current design has ~20 component layers for a blade. This allows for “standalone module” testing, but at a material price

• Reduce # of thermal (adhesive) interfaces = less material and thermal impedance

• Need method to evaluate bump bond connections before next assembly step = probe testing BBMs before module assembly


21-22 May 200818

Upgrade Integrated Module Concept

• Flip chip modules mounted directly on high heat transfer/stiff material (ex. pyrolytic graphite).

• Wirebond connections from ROCs to high density interconnect/flexreadout cables through holes in rigid support / heat spreader

• Leaves pixel sensors uncovered for scanning with pulsed laser• Flip chip modules REMOVABLE, leaving multilayer interconnect bus

intact for replacement modules.


21-22 May 200819

CO2 Cooling• A CO2 cooling system was designed and constructed for the

VELO detector in LHCb and will run in conditions (silicon detector, high radiation) that are comparable to CMS and ATLAS conditions

• CO2 properties are good for silicon detector applications– Low viscosity and low density difference between liquid and vapor is

ideal for micro channels (d<2.5mm)– Ideal for serial cooling of many distributed heat sources– High system pressure makes sensitivity to pressure drops relatively

small– High pressure (up to 100 bar) no problem for micro channels– Radiation hard– Environment friendly, ideal for test set-ups– Optimal operation temperature range (-40°C to +20°C)

• “No showstoppers” foreseen using existing CMS pipes for CO2 cooling, but modifications will have to be made to the LHCb CO2 system to reduce the pressure for CMS pipes

• CO2 cooling may be the best coolant for any upgrade in the CMS and ATLAS inner detectors


21-22 May 200820

CO2 Cooling

Consider cooling tube at edge of panel with pixel modules mounted on heat spreader

2D FEA model of the FPIX blade heat sink coolant temp of -15C

L. Cremaldi, U. Miss.

Small diameter (1mm) pipes for CO2 cooling:• much less mass ~1/10• small area for heat transfer - have to route enough tubes for

sufficient thermal contact with pixel modules• lends to design similar to current FPIX - flat substrates for module

support and tubing loops need for material budget optimization -- passive high thermal

conductive panels vs. routing small diam. CO2 cooling tubes to heat sources


21-22 May 200821

Activity Matrix

Institution - topic Activity 1 Activity 2 Activity 3

Purdue – integrated module assembly

Evaluate adhesives and alternatives for dense multilayer interconnects

Develop prototype integrated module assembly robot

Test prototype integrated modules, evaluation of radiation hardness

Purdue – integrated support/cooling

Evaluate materials for cooling tubes with a minimum of connections

Fermilab – integrated support/cooling

Study carbon fiber and oriented graphite composites using FEA models and prototypes

Construction of a CO2 cooling system for lab bench testing

Measure cooling performance of prototype integrated module-on-support structures

Develop conceptual design that integrates CO2 cooling with overall detector design


21-22 May 200822

ScheduleYear FY 2008 FY 2009 FY

2010

Task list Q1 Q2 Q3 Q4 Q1

X

X

X

X

X

X

X

Q2 Q3 Q4

Conceptual design cooling/support structure (with revision)

X

Q1

X

X

X

X

X

CO2 cooling study - construction of a CO2 cooling system

X

X

X

X

Study composites with high thermal conductivity for fabrication of low mass support frame

X X

X

X

X

X

Investigate cooling channel material alternatives

X

X

X

X

X X X

Evaluate design and materials support/cooling structure (FEA)

Measure cooling performance of prototype integrated module-on-support structures

Evaluate adhesives and dense multilayer interconnects

X X X

Develop integrated module assembly robot

Test prototype integrated modules (mech, thermal, electrical properties with radiation)


21-22 May 200823

FY08-FY09 R&D ScheduleEvaluate adhesives:

Q2-Q3 FY08: Identify candidate adhesives (In progress)Q3-Q4 FY08: Test radiation hardness by measuring mechanical strength of samples at low operating temperature before and after irradiation with fluence up to 10E16 protons/sq cm Q4 FY08: Measure thermal conductivity of module samples assembled with candidate adhesives Q1 FY09: Evaluate suitability for semi-automated assembly and integrated module repair

Milestone Q1 FY09: Identify reliable, repeatable and reworkable adhesive(s) (robust for SLHC conditions) for pixel integrated module assembly and rework. Evaluate dense multilayer interconnects:

Q2-Q3 FY08: Identify candidate state-of-the-art substrates (In progress)Q3-Q4 FY08: design and fabricate sample substrate multilayer interconnectsQ1 FY09: test radiation hardness, electrical performance, and wirebondability

Milestone Q1 FY09: Identify optimal product(s) for dense multilayer interconnects.Conceptual Design of cooling/support structure:

Q4 FY08: identify requirements for CO2 cooling and low mass/stiff/high thermal conductive supportQ4 FY08-Q1 FY09: CAD design of conceptual modelQ2-Q3 FY09: perform mech. and thermal FEA of CAD model – feedback to design revisions

Milestone Q3 FY09: optimal conceptual design for replacement/upgrade FPIX detector, ready to proceed to preliminary design for TDR


21-22 May 200824

FY08-FY09 R&D Schedule (cont.)Study composites with high thermal conductivity

Q3 FY08: identify candidate low mass/stiff/high thermal conductive materialsQ4 FY08-Q1 FY09: fabricate sample structuresQ1-Q2 FY09: test mechanical and thermal performance of samples

Milestone Q2 FY09: Identify optimal composite materials for low mass support structure

FY09-FY10 R&D ScheduleCO2 cooling study

Q1-Q2: investigate cooling channel material alternativesQ2-Q4: construction of CO2 cooling system for bench testing

Milestone Q4: reliable CO2 cooling system for testing of prototypes

Evaluate design and materials for support/cooling structure:Q1 FY09-Q1 FY10: perform FEA and test performance of integrated modules-on-supports

Milestone Q1 FY10: Identify optimal design and construction for low mass support/cooling structure.

Prototype integrated module assembly robot:Q1-Q2: design and fabricate tooling for prototype integrated module assemblyQ3-Q4: motorize and program assembly tooling

Milestone Q4: Operation of robot for reliable, repeatable assembly of integrated module prototypes.Test prototype integrated modules:

Q2-Q3: testing of mechanical prototype integrated modulesQ3-Q1 FY10: testing of electrically working prototype integrated modules

Milestone Q1 FY10: Demonstrate reliability of assembled prototype integrated modules.


21-22 May 200825

Relation to other WG and R&D activities• Our R&D plan to develop a lightweight and highly integrated

support structure using CO2 cooling for FPIX is a complement to the PSI group proposal for a phase 1 BPIX upgrade including: – a very lightweight structure with CO2 cooling– streamlined services and integrated modules– compatible with existing CMS services

• Our plan to develop a lightweight and highly integrated CO2 cooling/support structure for FPIX is a complement to the R&D at UCSB of an improved cooling scheme for an upgrade tracker

• We share an interest in CO2 cooling with ATLAS upgrade groups, particularly in the development of CO2 cooling for LHCb VELO by NIKHEF. It will be very useful to collaborate with these groups on the development of CO2 cooling for a low-mass Phase 1 pixel upgrade for CMS.


21-22 May 200826

SummaryPlan for upgrade mechanics R&D for next two years at

Purdue and Fermilab:• Addresses aspects of the current pixel detector for

improvement• Goals for upgrade mechanics• Tasks, schedule and milestones

In the next two years, we plan to be prepare an optimal design solution for structural support and cooling, and plan for assembly of an upgrade FPIX detector for Phase 1 in 2013, including:

• Requirements and specifications• Preliminary design• Structural and thermal analysis – performance predictions• Prototype assembly and testing

Preparation for Technical Design Review ~18 months from nowConstruction phase ~2.5 years from order of components (based on current Fpix experience).


21-22 May 200827

Responses to CMS Upgrade Management Board comments on SLHC Proposal 7.15

1. Reevaluate the timescale to deliver a sound conceptual design, including sufficient feedback from the evaluation phase.

Answer: we will make a conceptual design by Q1 FY09, and focus all development and analysis toward optimizing and completing the conceptual design by Q3 FY09.

2. As a great deal is to be accomplished in a short time, explain more about the linkages between the various tasks in the project.

Answer: we will proceed from an early overall detector conceptual design, to identification of requirements for module, support structure and cooling design, to parallel development of critical items that must be tested to meet the requirements, and feed back the development and test results to the overall conceptual design. A MSProject schedule is being made to clearly show the linkages between tasks and to track progress.

3. Explain the relation of this project to other pixel R&D in CMS and possibly ATLAS.

Answer: see slide 24.


21-22 May 200828

Backup slides


21-22 May 200829

Resolution of Flat Disk VS Turbine at 20o

Improved vertex resolution by factor of ~2 (raw estimation)


21-22 May 200830

Detector design ideas

• R&D will greatly depend on the detector design1. Keep 20° tilt angle of sensors2. Al cooling channels replaced by small diam. Cu-Ni (if high

pressure CO2 is used) or PEEK, LCP, or CF3. Consider edge cooling 4. Inner and outer Al support rings CF5. Use HighTC heat spreaders (TPG) and cool from outer

ring 6. Possibly small CO2 tubes running in grooves machined

in Be or C-C manifold around the edge

• A “flat disks” Fresnel design may also work (to reduce material), but loss in single hit resolution is troubling.

(with valuable input from lucien cremaldi, u. miss. simon ... documents... · wirebonds and >25...

Documents